An Interview of

Dr. Michael Merzenich – Neuromodulation and Neuroplasticity in Reading Related Brain Processes

Dr. Michael Merzenich is the Chair of Otolaryngology at the Keck Center for Integrative Neurosciences at the University of California at San Francisco. He is a scientist and educator, and founder of Scientific Learning Corporation and Posit Science Corporation, companies that develop therapeutic programs for the neurologically and psychiatrically impaired. Dr. Merzenich is an expert on 'brain plasticity' underlying the development of skills and abilities through experience and learning. Among his awards are the international Ipsen and Zülch Prizes, honoring his work in brain plasticity. Dr. Merzenich is a member of the National Academy of Sciences and is the author of the widely read "On The Brain" blog. Additional bio info

The following interview with Dr. Michael Merzenich was video taped at the offices of Posit Science Corporation in San Francisco, California in July of 2004. Our interview with Dr. Merzenich ranged from how the brain processes bits of sound to how our emotions modulate the processing of complex cognitive tasks, particularly those involved in reading. Of particular importance is our conversation about the processing inefficiencies associated with 'fuzzy' sound representations and their parallels with the inefficiencies associated with processing letter sound confusions.

Note: Dr. Merzenich also appears in the following COTC Video Segments: Reading Matters -  Unnatural Confusion - Nature and Nurture - Sensitive Slopes - Emotion and Cognition - Language Foundations

Video: Part 2 of 7: Signal-to-Noise & Language - Neuroplasticity & Sound
The entire interview is available for online viewing:
click here for the video index

Note: Remember to click on any word on this page to experience the next evolutionary step in technology supported reading.

The following transcript has not been edited for journal or magazine publication (see 'Interview Notes' for more details). Bold is used to emphasize our [Children of the Code] sense of the importance of what is being said and does not necessarily reflect gestures or tones of emphasis that occurred during the interview.

Personal Background:

David Boulton: Let’s start with a brief sketch on you, what you’re doing and why you’re doing it.

Dr. Michael Merzenich: My name is Michael Merzenich and I’m a Professor at the University of California at San Francisco where I have an integrative Neuroscience Laboratory. I am interested in complex processes of the brain related to learning, remembering and thinking and I’ve been very interested in the plastic processes that contribute to child development and to adult and life long learning. I’ve been specifically interested in those processes that contribute to variations in human ability and to providing an explanation for why one child might be clumsy and another agile, and one child might be a facile reader or facile in their language abilities or communication arts and another child might struggle to read, might struggle to develop language, might struggle in their communication arts and so forth. Finally, I’ve been interested in developing strategies to use this incredibly powerful process of brain plasticity, the processes that underlie the evolution of functionality in a brain through learning and experience, to command them to control them, to master them, you could say; to drive correction in a brain so far that that’s possible in a child that is impaired or an adult that is impaired, an individual that’s struggling, or an individual that has difficulties. So I’ve been very motivated to understand these processes on a level in which there could be some level of driving true correction – true neurological correction. I’m not talking about necessarily substituting function – one function for another around processes that aren’t correctable – I’m talking about, in so far as it’s possible, to drive true neurological correction in the brains of individuals that are struggling and experiencing difficulty. That’s primarily what I am motivated to do.

Neurological Progression and Correction:

David Boulton: You say neurological correction.

Dr. Michael Merzenich: Right.

David Boulton: Do you mean a contrived environment that allows a person to learn differently in relation to a challenge in a way that will bring about some compensational change in their neurology? Could you describe that?

Dr. Michael Merzenich: Every brain evolves in its ability to operate as a processor, as a receiver, as a user of information. One of the main sources of explanation for variations of human ability is how inherited weakness or differences of environmental experience contribute to differences in how the brain passes through those progressions; for example, the development of the language abilities of a child or the movement abilities of a child. If I look at any given child I see a very idiosyncratic individualized progression, as the child goes from the first wiggles to rolling over and sitting up and crawling, standing, walking, running, and jogging – ultimately they’re able to control very complex sequential behaviors, multiple movements, and multiple behaviors simultaneously. If you look at any individual child, they go through that progression in a virtually idiosyncratic way – no two alike – each child goes through its little progression in which you can say skill is built on skill, ability built on ability.

Ultimately, you could say if a child went through that progression, maybe with inherited weakness, and ultimately came to a point of real limitations in the ability to control movement, let’s say as a 5 year old, a 6 year old, as a 10 year old, to what extent could I, on the basis of understanding that progression that the child has gone through, could I know how to correct in so far as correction was possible? How could I intervene to correct it right at its core so that I, in a sense, make up for the missing pieces, for the missing deficits, so that I can evolve in that child the ability to control elaborate and agile movement? And I think in many instances that is the case and that is certainly the target of what we try to do.

David Boulton: So in some ways analogous to the somatic rehabilitation therapies that are out there?

Dr. Michael Merzenich: You could say that it is but a difference is that it is really based upon – in so far as we really understand them – the neurological principles, the neurological progressions that we understand underlie these evolving skills.

David Boulton: So the intervention would be more ecological and more targeted to the kind of differentiation you are trying to get the brain to do?

Dr. Michael Merzenich: It’s very specifically neuroscience based in consideration of how different progressions lead to different outcomes. It’s very deeply concerned with the different scenarios that play out in different individual brains, that is to say in different individual children. And, basically, how you would go back into such a brain and drive a true correction.

David Boulton: So does this require you to develop a number of unique assessments to be able to identify where these intervention points might be?

Dr. Michael Merzenich: Well, in some instances, in some classes of problems that children have, there are very strong commonalities, that is to say you could look at it statistically. Every child that has a language impairment does not have it for identical reasons. In the surface detail, it will have many different expressions. On the other hand, the faults that underlie a defective language representation in the brain that limits operations of memory and thinking and usage of language have much in common in the great majority of children that have such an impairment. So there are fundamental aspects that are seen over and over again in children that have such impairments. And those can be attacked and be addressed.

So on one level, there are these great common tools. On another level, there are all kinds of details that have to be worked on and worked with to drive those specific improvement and corrections that apply more individually, more idiosyncratically. So the most powerful approach is a combination, you could say, of the powerful general tool, hypothetically powerful general tool, and a whole variety of very specific targeted tools that basically fill in the details.

Signal-to-Noise:

David Boulton: Let’s get specific about the oral language processing you were just describing. You said that underneath the surface areas you might see in terms of the details of an individual child, there are certain commonly implicate things being developed. What are some of those? And bridging off of that, how do those differ and how are they similar to the challenge of learning to read?

Dr. Michael Merzenich: Well, let me give you a very simple example. If I look at the brain as a self-organizing machine and I ask what is the most common expected fault in the self-organization of its processing abilities, a very common general fault in its self-organization is the signal to noise conditions in which it organizes it basic language representations. And the reason for that is that there are many inherited faults that would contribute to degrading the capacities of the brain as a signal receiver or that would increase the conditions of noise that occur in an active brain.

You could say that in any such brain the extent to which those inherited faults contribute to such weakness the brain’s language is not going to be clear English. It’s not going to be English or Spanish, it’s going to be noisy English or noisy Spanish. That is a very common expected fault. There must be and we know there are variations of signal to noise conditions in the developing brain. And we know that the extent to which the processes in the brain are noisy as the native language is acquired, the speech model that will apply to setting up the processor and developing a processor and all its elaboration will be constructed on the basis of that defective model. So that’s an example of an inherited fault or class of an inherited fault that, in a sense, would come down to a final common pathway weakness that we could expect to target with a specific class of training tools.

Examples of Signal-to-Noise Effects on Language Acquisition:

Dr. Michael Merzenich: When you look at children that have language impairments, you commonly see that their processor has been adapted or adjusted just as you would expect to see it adapted or adjusted if the language model were noisy. Look in the brain of a child at six months of age, and look at how it responds to rapidly successive stimuli, or look at how it responds specifically to phonemic stimuli as it’s creating its selective specific phonemic representations that characterizes native language and you see it is defective in exactly the right ways. It misses those pieces of the speech and it has adjusted the processor just as you’d expect it to be adjusted if the problem is a problem of signal to noise and representation of the language.

For a second example, create a computational model of the self-organizing brain and say, ‘What would it be in the self-organizing brain that would contribute to the defects you would see in the language operations of the child?’ How you create that model is by simply adding noise by making the speech model noisy.

A third example is to look in an animal with a history of a noisy brain and then see how it adjusts its processors, its processing machinery, how it sets its time constants, its space constants as it processes complex acoustic information. It does it exactly the way a child does it in ways that account for its defective speech listening.

So there’s actually very powerful evidence that, at least in many children slow to learn language, the problem is a true delay in the development in the processing of their native language that leaves them with a defective language and their process is actually idealized not for English or Spanish, but for noisy English or noisy Spanish. It turns out that that defect, that limitation, is actually powerfully addressable plastically at any point later in life. You can see that in an animal. You can see that in a kid.

So, from such examples of the failure modes of the self-organizing brain or machine and the specific changes that occur in the brains of children that are struggling, you begin to see how and what accounts for their struggle. You begin to see what accounts specifically for their difficulties, but you also have great insights into how to potentially correct it. That’s basically the kind of strategy we apply in looking for these specific scenarios that may account for substantial general weakness in children that are struggling. Did that make sense at all?

David Boulton: Yes, it did to me.

Neuroplasticity and Sound Representations:

Dr. Michael Merzenich: You can actually create a speech model and measure the processing time in a child or an adult with this history and say if I represent speech by a processor that operates with this sampling rate, how clearly, how intelligible is it to a normal hearing individual? For the majority of these children it is very poorly intelligible, it has very limited intelligibility. That’s another way of saying they’re not extracting information from the speech input stream in the way a normal child is.

Now it turns out that dimension, we’ve shown in many experiments, is plastic. I can take you, or a monkey, or a rat, or a child with such a problem, or a normal child, and improve it. In fact, I cannot only improve it, I can degrade it. I can change it positively or negatively – it’s plastic. I can make such distinctions about stimuli very important to the brain by controlling the behavioral context and generally I can drive improvements in it. When I do that, the speech is represented in a clearer way. It’s represented to the brain in a more salient form. Now you can look in the brain and you see the time constant is more normal. Now you look in the brain you see that the way the brain represents information is more powerful. It’s representing it more coherently sort of event by event as the intra-syllabic information flows by. It’s represented in higher fidelity and every operation the brain makes with it is enabled by that.

In other words, in terms of the memory system, if the information flowing into the brain is represented in a non-salient way, it means that the capacity of the brain to encode and record it is degraded. If I refine it, if I strengthen it, I strengthen the capacity of the brain to record it reliably as a memory. So there are whole series of downstream, important, crucial consequences that occur from having a clear signal in.

This really goes back to the basic arguments that were made at the early era of the study of dyslexia, by Isabelle Lieberman and others, who argued that the fundamental problem is the degraded way in which a person’s speech represents this crucial phonemic information. It is representing it in a fundamentally degraded way but it is subject to refinement. You could say the engrams, or the story representations of information that are the basis of usage, are in a fuzzier, degraded form. It’s an old idea but it is exactly what we believe is happening because the speech model in such a brain is just not up to par in the early life of a child. It’s noisy and degraded. A brain that represents speech based upon a muffled model generates a processor idealized for processing muffled speech.

David Boulton: Good. Sometimes we use the analogy of a player piano, and that the code that we are reading with is analogous to the code on the scroll used for a player piano.

Ultimately it comes down to: Do we have the right number of keys? Is each key making its sound correctly? Can the keys be played together in cords? Is the piano out of tune? If the keys the code is assuming will be played aren’t working right, the sound ends up muddled instead of playing a clear comprehensible stream that reflects the code.

Dr. Michael Merzenich: Yeah, that’s a very nice analogy. Psychologists and neuroscientists have thought about these problems for a long time and they’ve isolated the sort of sub problems that apply to a dyslexic or language impaired child. For example, they might say in relation to a player piano analogy, we will define this as a problem. A child might have a problem with syntax, that is to say reliably saying what internal representation of sound goes with a sound in an incoming speech stream and making all the correct sort of online assumptions and relationships with respect to it.

All of those relationships, all of the ways in which you use information receiving is all dependent upon signal qualityIt is all dependent on how sharply, how crisply, how strongly, how effectively, how reliably the brain represents little pieces of sound flowing in. If you have a fault at the bottom of this chain then you have a problem with everything. And that’s fundamentally what the problem is.

I want to say one thing farther about this. In a sense, another way that people confuse themselves about this is that they imagine that there can be a fault at the bottom of the chain without a fault everywhere or that they imagine there could be a fault at the top of the chain without a fault at the bottom. This is wrong because no part, in fact, in a language system operates in isolation of another part. Faults are shared, right? If you have a fault at the top, you have a fault at the bottom and vice versa.

David Boulton: They are co-implicated. 

Dr. Michael Merzenich: They are co-implicated. There is a common argument that there can be truly a deficit in something like your ability to know, to devolve a word into its sound parts and to recognize that it has those parts. But in fact, a brain that does that does that reliably and normally has adequate processing across its operational system by necessity. A brain that fails to do that reliably is expressing a problem or failure of operation of the brain across its operational system. They are complexly interconnected top to bottom and bottom to top and you cannot really think about them operating in isolation. It’s fallacy.

The Genetic - Environmental Spectrum:

David Boulton: Let’s talk about how this relates to reading, which is our focus. Let’s talk about the difference between a genetically driven variation in this signal to noise distinction process and the quality of the sound scape in the environment the child is developing in.

Dr. Michael Merzenich: Right. Well, first of all, I want to say that every child that has a reading impairment, does not have an identical genetic fault.

David Boulton: I was talking about, for right now, the type of sound distinctions you were referring to that are at the core of all language disabilities and which form this kind of nucleus that gets ‘played’ by the reading processor. Put the reading aside, just down at the sound level, there’s a variation in how well children process the sounds. The sound processing is the core piece we want to understand before we go on to talk about reading. Relative to that variation, there is some of it that’s a genetic/structural developmental difference, then there is some of it that results from an impoverished environment – insufficient signal distinctions going on in their interactions with their parents and in their surroundings. What is your sense of that spectrum?

Dr. Michael Merzenich: The development of human ability represents a complex interplay between inherited resources and environmental influences. Our abilities absolutely spring out of the self-organization of our capacities as a product of our environment given our inherited resourcesSo both of them are absolutely crucial. You could say we have this broad category that we call normal development, which has a lot of slack, a lot of latitude in it. On the other hand we have the capacity to get the most out of a brain which means we can get more out of brains if the environment is enriched and positive and has greater possibilities expressed.

There’s been sort of a religious notion that we’re stuck with our inherited resources, and to some extent, of course we are. It represents a source of limits to every one of us – every individual human being. But at the same time, we have a tremendous capacity to modulate the outcomes of this given, our inherited resources, as a function of our individual experiences.

One of the complications of that is there is a very poor understanding of what good is when we say what we need to do for a kid, for the kid to advance in the most positive way and have the most general positive benefit from it. So we’re very confused about what exactly it is that the brain needs to get the most out of it, to make it the best at communicating, and ultimately, become the best reader. But given that aside, absolutely environment makes an enormous difference.

So for example, other scientists have taken rats and demonstrated that you can, in a wide variety of ways, enrich their environments. So you can train a rat with all kinds of complex sounds in music and other things occurring in their life that are important to them and you can demonstrate that the processing of sound in their brain is much more elaborate, much more powerful, much more detailed processing, much more complicated processing, much more selective.

I can raise a rat under conditions in which particular classes of sound that have particular features of complication are really important to the rat and the brain of the rat will be entirely different than a brain of a rat in which that is not the case. I can raise a rat in which I make the rat an expert in some particular domain of sound. Let’s say I make the rat an expert at making fine distinctions about subtle changes in the loudness of sound and I look into the brain of the rat and I see the machinery that relates to representing details of the loudness of sounds are beautifully over-expressed, beautifully elaborated in the rat and so forth. So in all kinds of detailed ways you can change the outcome as a function of the processing capacities of the brain as a function of what the brain does. What its history is. What it has learned about. What its progressions are.

Brain’s Listening Dominated by Language:

Dr. Michael Merzenich: Now in our case, in the human case, the listening experience of the brain is overwhelmingly dominated by language experience. It’s a powerful, language is an incredibly powerful, experential source in the average kid.

A Dutch scientist counted the words that his child had heard by the time he was nine months of age and he found the child has been in the presence of about four and a half million words, of which about half a million had been spoken to the child. The child had actually had half a million words directed to him, of which about fifty thousand were produced in highly exaggerated form, in parent-ese, where the parent is talking in baby talk to the baby so the child maybe can follow the speech a little more clearly. That’s before the child puts the meaning to the first word. That’s massive experential exposure.

Now you and I know families at which that same moment in time, a child may have not been in the presence, except maybe on the television, of fifty thousand words. Does that make a difference? Very probably. It very probably makes a difference and the more the parents are interacting with a child, in a sense, the more that the received speech means to the child, probably the greater value it has as a training signal. It will make a difference. Sound experience in a child will make a difference. The way the processor of the brain will be set up will be largely dominated to specialize for the native language the child is exposed to because that sound exposure is so massive and so special in the history of learning in the child.

David Boulton: It is so central to everything they’re listening for in order to pull themselves into their world.

Dr. Michael Merzenich: Absolutely, and it becomes so quickly massively rewarding. I mean it has this wonderful aspect that it becomes a main source of reward for the child, moment by moment in time in life.

David Boulton: That ties in with the anthropological evidence that the evolution of the structures of our throat, jaw, tongue, and teeth have been selected to conform to the demands of articulating spoken language.

Dr. Michael Merzenich: Well, one of the things that happens, of course, is that it’s massively exercised. And because it is so massively exercised it’s also in every individual’s life history. It is incredibly refined and specialized for that specific purpose. You’re not just receiving. By the time you’re nine months of age you’re not talking very much, you’re mostly talking in the phonemic and syllabic kind of ways. But pretty soon you’re producing speech on a massive schedule and that also represents heavy motor exercise and sensory exercise for your face and pharynx and tongue and so forth.

Language and Literacy:

David Boulton: All of which are participating in a somatic feedback bath during the process of differentiation, which is part of what makes this level significantly different than the subsequent processing involved in literacy.

Dr. Michael Merzenich: Absolutely. Of course when you get into reading then you enter another great epoch of intense exercise and learning that ultimately becomes a massively practiced skill, almost unprecedented in human history beyond language. It has almost no precedent.

David Boulton: You say almost. What’s even close?

Dr. Michael Merzenich: Well, language itself of course, you could say in a sense, surpasses it. But beyond language – nothing. I can imagine other things you do on a very heavy daily schedule that are pretty substantially perfected but …

David Boulton: According to an external set of instructions that are….

Dr. Michael Merzenich: Beyond language and beyond aspects of movement, nothing.

An Incredibly Modern Invention:

Dr. Michael Merzenich: One of the interesting things about reading is that it’s an incredibly modern invention. I mean, in a sense, reading was invented by Gutenberg. I once heard Professor Gardner at Harvard make such a statement and I’ve loved it ever since. You could say reading was invented by the Swedish kings, who for the first time required that every child in the kingdom must learn to read and who left the local preacher, the local minister responsible for teaching them. And that’s a 17th century event. So in fact, this incredibly heavy exercise that we’re all engaged in, that involves such intensive learning is based on a relatively recent event.

I might say there is another thing in the modern life that approaches it in many children and that’s computer usage. The facile use of the computer, mouse and keyboard in ways that, you could say, are beyond reading, in communication, with a machine.

David Boulton: There are few who learn to use a computer if they are fundamentally illiterate, unable to read or otherwise communicate complexly.

Dr. Michael Merzenich: Well, that’s true but there is a class of individuals that are absolutely illiterate and can’t read and are in fact even mute, that are very facile computer users. I mean we’ve done studies on autistic children. One of the things we’ve seen in autistic children who are mute and can’t tell you anything about anything, is that if you have in your computer training program information about how to manipulate the program by punching function keys, they quickly discover those and the next thing you know they are wildly racing around the computer. You could find very few normal children that have such facility. So, you know it is not, it does not derive directly from reading.

David Boulton: No, but let’s put it another way; somebody that could physiologically be able to read and who can speak, who can hear language – if they attempt to learn to read and don’t make it, they have certain problems that are probably going to make it difficult to get good at word processing.

Dr. Michael Merzenich: Absolutely. It’s a great source of limitation to them in many ways. But it doesn’t mean that their reasoning abilities aren’t powerful or effective on the short term. On the long term, as we know one of the tragedies about having these impairments endure is that ultimately the measures of your cognitive abilities tend to move down in the direction of your reading abilities. It represents a real source of limitation and frustration in the child that hopefully can be overcome. It is very important to overcome it if it’s possible. That doesn’t mean that every smart person is an effective reader. I’ve met and you’ve met a number of truly extraordinary people that can’t read a lick and so it’s not strictly correlated with intelligence as we know and we see these incredible exceptions. 

Emotional Dimensions:

David Boulton: One of the things common in most of your exceptions is they had enough strength of self to not get hurt by not being able to read.

Dr. Michael Merzenich: Right. Well, to say they’re not getting as hurt is one way to say it.

David Boulton: Psychologically hurt. They don’t get so ashamed.

Dr. Michael Merzenich: I once asked a very prominent American, very successful individual, one of the hundred most wealthiest Americans, why, given his great individual success, he felt investing in dyslexia related research was so important? Why did he cared so much? It was clear immediately when the question was asked, why he cared so much. He still felt the pain of self-identifying as a kid as a failure and of perhaps that self-identification more than his good parents was still carried as a burden by him later in life. (More “shame stories”)

David Boulton: He clearly triumphed over it.

Dr. Michael Merzenich: He clearly triumphed over it, absolutely.

David Boulton: Whereas, you know, there are many millions who don’t turn out nearly as well.

Dr. Michael Merzenich: The prisons, you know the world of failure is full of them as we know – people who never recover their self-esteem from those early failures.

The Effect of Affect on Cognitive Processing:

David Boulton: Let’s talk briefly about that affective dimension. In your work as a neuroscientist are you able to see the effect of affect on cognitive processing?

Dr. Michael Merzenich: Absolutely. It is a main thing that we study and as you know there’s a long history of its study in experimental psychology. Experimental psychologists began more than a hundred years ago to ask what it is. What factors control learning? That basically modulate it? That make its effect affective? That make it stronger? That vivify it? This has been a great interest in psychology and a great interest in neuroscience ever since. We know, for example, that if you learn under the right emotional conditions, you never forget. Learning is very affective.

Take animal experiments as a very simple sort of example. If you give a rat something it really desires or something it really hates, it doesn’t take very many trials for the animal to learn a behavior. On the other hand, if you give it something that’s just a little bit rewarding, or just a little bit punishing, it can take a thousand practice tries or ten thousand practice tries to get that point. So there are these modulators of learning that apply in every older brain.

One of the interesting things is that we don’t come into the world with this. In the very first part of life, in what’s called the Critical Period, the learning machinery is on continuously. It doesn’t really care so much about how important the input is to the brain – it has got to change and it changes on the basis of being in the world. It is kind of a set up time where the learning machinery, basically this modulatory control, is just turned on. The learning switch is just saving everything. What happens is the brain evolves, at the end of the critical period, the modulatory control comes into play. Now you learn most effectively, overwhelmingly, when you’re paying attention and as a function of the rewards and punishments that occur in the behavior and how bright and alert you are in the behavior. All of those things are critical contributors to how effectively we induce change in learning. We’ve studied this in a variety of ways and many others, hundreds of other scientists have studied this.

One way that we’ve studied it is to control the chemicals that modulate these processes one at a time and what we’ve been very interested in is what they contribute, process by process, to the changes that you witness in the brain as you learn. So for example, you can control the chemical that is primarily associated with a good outcome or reward, that’s dopamine, and if you control it selectively you generate one set of changes in the brain as you learn. Or you control a chemical that is primarily associated with learning under conditions of close attention, that’s acetocoline, which generates another set of changes in the brain and so forth. So when you learn you actually have nuances of change that are contributed by different aspects of the emotional state when learning.

Now let’s take a really big event like the day Kennedy was shot. Where were you? What happened? Most people that lived in that period remember that. Man first on the moon was thirty-five years ago. Where were you? Remember that? Most people will remember some things about that, maybe the flag on the moon, maybe the first step, maybe the first phrase. I do and I will never forget it, and because all of those things were in play, all of those neuromodulators were in play I was paying close attention. It was a highly rewarding experience. I was alert as a mind, as a brain could possibly be as a young man at that point. I’ll say that in any such experience, in learning, in teaching, in the classroom, in anything you do in the brain, the most effective way to drive change is to have those conditions right and to control them. 

David Boulton: Have you ever encountered Sylvan Tomkins?

Dr. Michael Merzenich: No.

David Boulton: He was an affect-theorist dealing with ….

Dr. Michael Merzenich: I know the name.

David Boulton: One of the particular things that’s interesting, in addition to thinking of affect as this channel that cognition is happening inside of, that has these different biases: interest, fear, anger, so forth, is that this is itself a construction, a neurobiological construction, this carrier wave that is channeling attention…

Dr. Michael Merzenich: It’s absolutely a construction. I mean, an example of that is to look at the domain of rewards that would apply, let’s say of one individual versus another. What would one person find to be rewarding and what would another person find to be rewarding? If you look beyond the most primitive things you find fabulous elaboration that’s very individual, and you see big distortions, obviously, that can occur in one individual and don’t apply at all in another individual.

David Boulton: What makes something relevant?

Dr. Michael Merzenich: Right. And as you move through life, these become more and more idiosyncratically complicated. They’re a big part of what your brain signals to you as important and that’s actually being created in your brain as you move and as the brain organizes itself across life.

David Boulton: One of the things we’re interested in that connects here is shame. In particular, what’s happening, not just at the level where we say, ‘I feel ashamed of myself,’ but before that, preconsciously before that. In the admix of affect that’s powering and directing cognition when a child starts to go into shame, what’s happening to their cognitive processing bandwidth? Our sense is there’s this downward spiral that’s starting to kick in as the child starts to feel shame and the shame itself starts to become more and more occupying of attention and …

Dr. Michael Merzenich: Right.

David Boulton: Diminishing the cognitive bandwidth available for processing what they’re trying to do.

Dr. Michael Merzenich: It’s a great point, and it’s a wonderful thing to study. To some extent it is being studied but it’s a wonderful thing to try to develop a deeper understanding of. Let me just say that there is this sort of intuition that all of the change that occurs in these learning contexts is positive. That if I add emotion I learn better, right? Not right, that’s not correct.

Let me cite a simple example. If I learn under conditions of incredibly high stress, I’m extremely efficient, immediately reactive to the situation at hand, but I’m not learning worth a darn. Obviously, if I go into a period of depression and I come out of it, I can’t remember a darn thing that happened.

In an experiment, if I put an electrode in the brain of a rat or a mouse and I stimulate it each time it has an experience, a particular form of experience, so as to create a condition that would be just like very powerful activation of the brain systems that contribute to depression, what happens is that I generate an incredibly big positive exaggeration of that stimulus. It grows in the brain like a monster and everything else is degraded. So here I’m growing something that you could say is like a great fear or a great obsession, and its power in the brain is growing and growing. Everything else is degraded. So there is a negative side, a dark side to what’s happening as well. So neuromodulatory systems are not just contributing positive change.

I’ll give you another simple example. It was known since Pavlov, that if I follow a stimulus with a reward, first the bell then the meat, the dog salivates. If I ring the bell and the dog salivates, that’s called classical, or Pavlovian conditioning. Well, let’s flip it around. Let’s give the meat then the bell. What happens? The answer is a peculiar form of unlearning. What happens is that now, if I flip them around, and now I have the bell and then the meat, the dog can’t learn the relationship. I have a negative effect that is occurring.

When I look in the brain of a rat that has the same experience, I see a positive change occur when the bell precedes the meat, signaled by the release of this powerful neuromodulator. When I reverse them I see exactly the opposite. The ability of the bell to excite the brain is erased. I actually see a negative plastic consequence. So we think of this as being positive. It’s not all positive – we think of learning as positive, but it’s not all positive. 

Another way to think about it is learning is selective. It’s not just that you learn everything all the time – it’s selective. It’s a selective process. Once a brain scientist, Herbert Jaspers, who is a great scientist at the Research Institute in McGill in Montreal, (I might say he told me this when he was about 90 years old) said, “Never forget that when you’re trying to get across town, it’s not just a matter of what bus you get on, it’s all those busses that go by that you don’t get on.” He said it’s being able to sort out what’s important in a complex array of things that are happening and arriving in your brain that you have to make distinctions about. The brain processes that govern the development of our behaviors are selective. Some things they operate on positively, some things they’re suppressing and they’re operating on negatively.

Shame, Self-Esteem and Neuromodulation:

David Boulton: Have you done anything to actually isolate shame, develop a kind of signature of what shame looks like when it’s triggered in the brain?

Dr. Michael Merzenich: No. I’ve thought about it primarily from the point and read about it primarily from the point of, you could say, it’s obverse; which is self-esteem or positive self-awareness or self-healing and haven’t really thought about it as you raise the issue. It’s a wonderful thing to think about in the obverse. I’ve primarily thought and read about it in a sense in another thing that would reflect it and that’s the sort of the ongoing loss of self-esteem, of self-confidence, or in terms of self-doubting and what can be done about it.

David Boulton: I spent a lot of time in the self-esteem world before getting into the reading world. I’m interested in the reading world because of what it’s doing to self-esteem. Although I don’t like the term self-esteem, it’s got a lot of baggage with it.

Dr. Michael Merzenich: It does and it’s confusing.

David Boulton: I think of it as a buoyant absence of self-negativity rather than a positive accumulation….

Dr. Michael Merzenich: Right, exactly.You can have a boy that’s very self-confident, and yet in a sense, identifies himself in a broad domain as a failure and, in fact, his self-confidence is a kind of compensation. Actually, what I’ve really been interested in is the origins of bad conduct behavior because I think it would be a tremendously positive thing if we could understand it. Again, it relates to the development of emotional control and these complex reactions that occur in the brain that govern, in this overriding way, the general behavior of a child.

David Boulton: The sum of our view is this… We think that children are being overwhelmed with a form of confusion that is unnatural to them and they are learning to associate the feeling of that confusion with shame. We are all shame avoiders, escape artists; we don’t like to feel shame. So just as reading involves an assembly that’s faster than conscious, there’s a faster than conscious aversion to shame which is associated with feeling certain kinds of confusion, which in turn decapitates learning because learning involves extending through confusion.

Dr. Michael Merzenich: I love that description, although I don’t really understand its neurology. I love the description.

David Boulton: One of the things we’re trying to do is bring about some new explorations into the effects of negative affect triggering in the stream of cognitive processing. It seems to me it’s got to be dis-entraining.

Dr. Michael Merzenich: Right, right. I think it’d be a great leap if you could understand that. I mean, if you have a child that enters school and the child fails at school,which means that they’re experiencing, they’re reacting in these ways to it and if they’re misbehaving, they have about an even money chance later in life to commit a felony. I mean the simple fact is we have to figure out how to get to those kids more reliably and more completely and it doesn’t necessarily involve cultural re-education. The more deeply we understand the neurology the more we could understand how to drive true correction of it.

David Boulton: Unless the neurology work embraces and includes the interdynamics between the more mechanically cognitive processes and the affectual processes that are orienting, contextualizing, powering, the cognitive activity, I don’t see how we can get there.

Dr. Michael Merzenich: Well, I don’t know. I believe all these things are at some level accessible. Whether they could be, whether such training could be easily incorporated under conditions of something like a public school where an aspect of the training might include some level of social training, whether that’s really necessary or not, I don’t know.

David Boulton: Well, it’s also a matter of re-contextualizing the kinds of challenges that most trouble children so that they are less provoking of shame.

Dr. Michael Merzenich: Well, one of the ways we’ve tried to design the software that we’ve applied to children, the training programs we provide to children, is to absolutely ensure a high level of success in every kid and this is a big part of it. A big part of it is to find ways in the course of every kid in a significant part of their life and day to ensure that they succeed in things. I mean one of the magic things that has to happen to every kid is they have to find something, something that they’re good at and everybody acknowledges and everybody tells them and they tell themselves. As long as you have enough of those things happening in your young life you can accept the things someone else is better than you at.

David Boulton: And the danger there is that it can be that the thing that I feel good about myself is that I can beat up everybody around me, right?

Dr. Michael Merzenich: Right, I mean there is that. There has to be things outside of that realm.

David Boulton: All right. I think we’ve talked about many of the things we’re both most interested in. Is there anything that we’ve inspired in the course of our conversation, anything you’d like to touch on that we haven’t yet?

Dr. Michael  Merzenich: Only to say that, this is a great quest. There is great hope that we can progress in understanding the neurological origins of things like this and the crucial dimensions as you’ve expressed, in this relationship between the emotional side of life and the overriding behavioral things that relate to neuromodulatory control in emotion.

I mean when you talk about neuromodulators and you talk about the things that condition learning, you’re also talking about the overlay of emotional baggage that comes with them, that are expressed with them, but also derive from them. You also talked about the fact that each one of us is basically creating the complex conditions under which they are constructed and they are inhabiting our heads and controlling our behaviors in our life.

One of the crucial things is to understand this complicated interplay between these modulators and understand how they evolve – how plastic they are. That’s been studied almost not at all. People have studied the axis of neuromodulation. Dopamine cells and dopamine chemistry has been manipulated in hundreds of ways partly because people see that there’s profit in it – to manipulate it because it’s so clearly involved in so much malfunction and so much disease in psychiatry and neurology. But still almost nobody has studied in any detail, or in any intelligent way, how this capacity, how this control, how this incredibly differentiated nuances of emotional control, as they influence learning and behavior are actually created and developed in an individual way and in an individual person. There are high stakes in such understanding.

So, one of the things I think is evolving or will gradually evolve is a level of science in that area that will parallel the level of science in learning as it relates to the sort of dry side of learning, and that’s skill learning. As that evolves I think we’re going to generate more and more powerful ways to intervene in real human problems. 

David Boulton: That’s excellent and that’s being echoed in other circles. Russ Whitehurst realizes that it is a weakness in the whole assessment paradigms of education. Other scientists that we’re talking to are also recognizing that this is a huge lacuna in our overall understanding and that it’s too significant to leave out of the equation.

Dr. Michael Merzenich: Right, and not just for kids. It’s a big part of what we’re trying to do with older individuals. I mean with a seventy, eighty or ninety year old this machinery is now falling apart and you have to think about how you can reinvigorate it, revitalize it, re-enrich it. You know, we have to. People’s attitudes can be better. There are things that are rewarding to them and can be made stronger and more elaborate and they should be. It’s part of being healthier and happier in older age. And guess what emerges when you get older? Shame. What emerges is a lack of confidence. I’m now worried about whether I can succeed, whether I can keep my end up. I mean all of these things enter in again in spades.

David Boulton: That’s why I think it’s so important that we understand that there is a biological basis for the core affects and that shame seems to be, when we think about the human animal, a fundamental learning prompt, a really great thing. But if we become averse to the feeling of shame before it actually bubbles up, before we can think about it, it’s steering us away from what we need to learn about.

Dr. Michael Merzenich: Absolutely.

David Boulton: And it seems to be happening on a grand scale.

Dr. Michael Merzenich: Right, right.  It just basically shuts you down.

Code Ambiguity:

David Boulton: Let’s talk about the kind of ambiguity that’s involved in processing this code. It seems that in everything else that we do, like we talked about earlier, there’s this somatic feedback field, right? One system gets feedback from another system to kind of build connections with. As a child as I differentiate my motor controls, I can feel the gesturing of my face in articulating a certain sound. I can hear, after you’ve said it many times, the carrier wave in your affect on one level of the meaning and my differentiation of the word sound matching up with it. There’s a lot going on. When I get to this reading thing, these letters are place holders for a wide field of possible sound values that aren’t resolvable in direct relationship with these letters, but rather have to be disambiguated. I have to buffer up the context and work out their values almost as fast as I can possibly think.

Dr. Michael Merzenich: Right, the brain is incredibly good at making these almost real time correlations, and it can do that, does do that with incredible power within your every action. I mean you’ve just described evolving your sense of feeling you, as you control movement. Well, in fact, you never actually do that – hardly ever. I mean you can look at the operations in isolation but nobody told the brain generally that you’re looking and feeling in an operation, right? In fact, you’re doing both and maybe looking, feeling, talking and smelling or whatever, right? And actually the brain is creating all of its constructs like this around this sort of ongoing active correlation and it’s incredibly good at it.

So, your face, your anterior face and your tongue and your throat, are becoming incredibly specialized as you speak. People tend to think about it and say, ‘Well I’m speaking and as I’m speaking, I’m listening to myself and in some sense I’m correcting.’ Well, just paralyze the sensation in your face and try to talk. You know you can’t talk very clearly, right? The reason you can’t is because you have all of this unknown contribution and feedback that are doing all of that production that you’re not aware of. You’re getting lots of feedback that’s not listening at all.

In fact, it’s pretty crucial to you being able to make all of that sound roll out. And that’s all a tiny way of saying that when you do something like an act of reading, where you’re making this ongoing translation between what you see and the brain’s representation of the sound parts of words that you’re associating and then on a higher level pretty soon you’re looking at whole phrases, maybe even whole paragraphs, and you’re just whip, whipping it up, right? You’re making those correlative associations in a powerful way, an incredibly powerful way. But the machinery is made to do that. It’s made to make ongoing correlations of that information and to make these elaborate constructs of it.

David Boulton: The brain is made to do that? You mean it’s generally capable of making complex correlations? But this code that is used today is only a few hundred years old. We’ve talked to the people who are experts in its origin, in the 1400’s with King Henry the IV and so forth. When you look at it, there’s a thousand different ways, twenty-six letters make forty-four sounds depending on the context, which can sometimes take a whole sentence to determine what the sound value is.

Dr. Michael Merzenich: I think I said it was good at it, didn’t I?

David Boulton: Yes.

Dr. Michael Merzenich: You’re just saying that it’s really, really good at it, right? I agree. I concede.

David Boulton: It’s incredibly good at it and it’s an incredibly unnatural challenge.

Dr. Michael Merzenich: Right, right, I agree.

David Boulton: In that for some children, for reasons connected to the auditory processing that we were talking about before…

Dr. Michael Merzenich: Right, it depends on what you call an unnatural challenge. If you think of what I do in my daily life, almost everything that I do, that well-developed skill that I have did not necessarily exist, even a thousand years ago. Pick up a spoon. Were spoons are a recent invention? Grab that bottle and drink it. Who ever saw a bottle a thousand years ago?

David Boulton: Right, but the difference between manipulating a physical object, as difficult as it is, and learning to process this code is that I can feel this object; it’s a relatively stable object in my universe. But when I’m looking at a letter, especially a five year old looking at a letter…

Dr. Michael Merzenich: But you see, you’re bringing in some magic into it that I won’t bring in to it, and here’s where I’ll chop it out. It’s absolutely right that you have an incredible facile ability. So now I spent tens of millions, maybe several hundred million practice trials generating a facile representation of phoenemic composition of syllables sound parts of words that I’m now going to represent by letters and their syntax. I’ve got a magnificent representation of that. Now I’m coming into that with all those resources, I’ve got those incredible stored and reliable resources in which I’m making all of that ongoing analysis and interpretation. Now I don’t start reading just lickety split. I have to struggle. Reading is not automatic, I have to relate those letters to those sound parts.

Now let’s start that and let’s spend a year or two, hundreds of thousands of practice trials before I’m really facile at that at all, and now I know what rhymes; this rhymes and that rhymes; I know that this word has these sound parts. I’m creating that reference library and now I’m creating it and now I gradually have a million or so other practice trials and I’m gradually getting it involved in and elaborating it into my semantic dictionary where I’m getting into the real syntactical elaboration and complexities of the real language and I’m getting into the phrases and sentences. In other words, this is an incredibly powerful self-constructed…

David Boulton: Virtual reality experience.

Dr. Michael Merzenich: Right, I mean this is not something that just pops out of the brain. This is the product of massive, massive learning.

David Boulton: A hundred million of us are not doing this very well.

Dr. Michael Merzenich: Yeah, that’s right.

David Boulton: And the consequence is a lot of our lives are harmed.

Dr. Michael Merzenich: Absolutely.

David Boulton: And the kind of confusions associated with working these letters into…

Dr. Michael Merzenich: I think what we’re both saying is the same thing really. We’re saying that it’s a miracle that it ever happens. It’s an incredible human capacity. It’s hard won by incredible depth of practice. It’s very unsurprising that many people struggle with it. 

Not Much of A Fault:

Dr. Michael Merzenich: I mean, you wouldn’t have to have much of a fault in this machine operating with high speed in this incredible processing efficiency that’s required to begin to see somebody be a little slower at it or a lot slower at it. And so therefore, dyslexia, in a self-organizing machine of this nature, is an expected problem. It’s an expected weakness. It should apply very widely.

David Boulton: In addition to the underlying sound processing issues we talked about earlier, wouldn’t the trouble also be related to the confusion in correspondence between letters and sounds? That the more brain time it takes to resolve the ambiguity…

Dr. Michael Merzenich: Absolutely.

David Boulton: The greater the stress on the system to produce…

Dr. Michael Merzenich: Make the representation of the sound parts of words a little fuzzy and it’s going to slow down the ability. And as soon as you do you also have a decline in processing efficiency, the ability to hear something and rapidly translate it or to do anything with it has slowed down.

David Boulton: Doesn’t that same exact argument apply to the fuzziness between sounds and letters?

Dr. Michael Merzenich: Absolutely it does.

David Boulton: That’s a technological artifact not a naturally occurring sound scape variation. That’s my point.

Dr. Michael Merzenich: Yes, absolutely. Yes, it is in a sense a technological artifact. Absolutely.

Special thanks to Jennifer Ware for transcribing this interview.